Image processing method and systems

• US 20030053689A1
• Filed: 11/28/2001
• Published: 03/20/2003
• Est. Priority Date: 08/27/2001
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Watanabe, Masahiro, Fujimura, Koichi
• Granted Patent: US 6,947,594 B2
• Time in Patent Office: Days
• US Class Current: 382/167
• CPC Class Codes: G06T 5/75 Unsharp masking
• CPC Class Codes: H04N 1/58 Edge or detail enhancement; Noise or error suppression, e.g. colour misregistration correction H04N1/62 takes precedence
• CPC Class Codes: H04N 1/60 Colour correction or control H04N1/54 takes precedence
• CPC Class Codes: H04N 1/6027 Correction or control of colour gradation or colour contrast H04N1/6058 takes precedence
• CPC Class Codes: H04N 1/6077 Colour balance, e.g. colour cast correction
• CPC Class Codes: H04N 1/628 Memory colours, e.g. skin or sky

Abstract

This invention provides a novel image correcting technique for automatically carrying out a suitable image correction. An image processing program of the invention includes a correction front-end unit for carrying out a color balance correction, a range correction, a main portion estimation processing, and a tone correction, a statistical information calculation unit for generating a color saturation reference value and a contour reference value as data expressing the preference of an operator by using an output of the correction front-end unit and a manually corrected image, and a correction back-end unit for carrying out a color saturation correction processing using the color saturation reference value stored in a reference value DB and a contour emphasis processing using the contour reference value. A processing result of the correction back-end unit is stored as an output image into an image storage DB.

Claims

1. A method for correcting a color fog for an input image, said method comprising the steps of:

calculating a statistical amount with respect to color saturation components of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a hue given as a correction reference value of said color fog, said color saturation being weighted by values of lightness components of said at least one group of pixels; and

correcting said color fog by using said correction reference value with respect to each pixel of said input image.

2. The method as set forth in claim 1, wherein said reference hue region is a hue region in which most pixels of said input image are included.

3. The method as set forth in claim 1, wherein

each of said plurality of hue regions is divided into a first region of a high lightness region and a second region of a low lightness region on the basis of a lightness value of a color having maximum color saturation in each said hue region, and

said reference hue regions are a region among said first regions of all said hue regions, in which most pixels of said input image are included, and a region among said second regions of all said hue regions, in which most pixels of the input image are included.

4. The method as set forth in claim 1, further comprising a step of calculating a statistical amount with respect to hue of said at least one group of pixels included in said reference hue region as a hue reference value of said color fog, and

said correcting step comprises a step of adjusting said correction reference value on the basis of said hue reference value.

5. The method as set forth in claim 4, wherein, in said adjusting step, said correction reference value is adjusted in accordance with a cosine value of a difference between a hue value of each pixel of said input image and said hue reference value.

6. The method as set forth in claim 1, wherein the weighting is performed such that as a difference with a lightness value of a color having maximum color saturation in said reference hue region becomes large, the weight becomes heavy.

7. A method for correcting a color fog for an input image, said method comprising the steps of:

calculating a statistical amount with respect to hue values of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a range of hue as a hue reference value of said color fog;

calculating a statistical amount with respect to color saturation components of said at least one group of pixels included in said reference hue region in accordance with a predetermined method as a correction reference value of said color fog; and

correcting said color fog with respect to each pixel of said input image by using said correction reference value adjusted by using said hue reference value.

8. The method as set forth in claim 7, wherein, in said correcting step, said correction reference value is adjusted in accordance with a cosine value of a difference between a hue value of each pixel of said input image and said hue reference value.

9. A method for performing a range correction for an input image, said method comprising the steps of:

detecting a highlight pixel having highest lightness and a shadow pixel having lowest lightness in pixels included in said input image;

converting values of respective color components of said highlight pixel in accordance with a specified maximum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said highlight pixel and minimum values which said respective color components can take;

converting values of respective color components of said shadow pixel in accordance with a specified minimum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said shadow pixel and maximum values which said respective color components can take; and

with respect to each of said color components, linearly converting a value of the color component of each pixel of said input image, which is contained in a range from a value of the color component for said shadow pixel before conversion to a value of the color component for said highlight pixel before conversion, into a value of the color component in a range from a value of the color component for said shadow pixel after conversion to a value of the color component for said highlight pixel after conversion.

10. The method as set forth in claim 9, wherein, in said step of converting values of respective color components of said highlight pixel, a maximum value among values of the color components of said highlight pixel is converted to the specified maximum gradation value, and

in said step of converting values of respective color components of said shadow pixel, a minimum value among values of the color components of said shadow pixel is converted to the specified minimum gradation value.

11. A method for specifying a noticeable portion in an input image, said method comprising the steps of:

dividing said input image into a plurality of regions;

calculating a rate of a human skin pixel by counting said human skin pixel, which is a pixel satisfying a previously set human skin condition, with respect to each of said plurality of regions, and calculating an average and a standard deviation of said rate of said human skin pixel for said plurality of regions;

judging whether or not a region including a portion estimated to be a person on the basis of said average of said rate of said human skin pixel exists; and

if it is judged that the region including the portion estimated to be the person exists, setting an importance level expressing a most noticeable portion to the region including the portion estimated to be the person, and setting an importance level lower than said importance level expressing the most noticeable portion to a region including a portion estimated to be something other than the person on the basis of a value of the standard deviation.

12. The method as set forth in claim 11, further comprising the steps of:

with respect to each of said plurality of regions, calculating average lightness, a rate of a sky pixel, which is a pixel satisfying a previously set sky condition, and a rate of a cloud pixel, which is a pixel satisfying a previously set cloud condition, by counting a number of said sky pixel and a number of said cloud pixel, and calculating an average value and a standard deviation of said average lightness for the plurality of regions;

judging whether said input image is in a backlight state on the basis of at least one of said average lightness, said average value and said standard deviation of said average lightness, and said rate of said sky pixel and said rate of said cloud pixel; and

if said input image is judged to be in said backlight state, setting an importance level expressing a most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said average lightness and said average value of said average lightness, or a portion which is not the dark portion due to the backlight but is a portion estimated to be something other than the cloud and the sky on the basis of said rate of said sky pixel and said rate of said cloud pixel, and setting an importance level lower than said importance level expressing the most noticeable portion to other regions.

13. The method as set forth in claim 12, further comprising the steps of:

if said input image is not judged to be an image including the region including the portion estimated to be the person and an image in said backlight state, setting the importance level expressing the most noticeable portion to the respective regions.

14. The method as set forth in claim 13, further comprising the steps of:

by weighting said average lightness and said standard deviation of the lightness, which are calculated with respect to each of said plurality of regions, by the importance level set for each of said plurality of regions, calculating a weighted average value of said average lightness and a weighted standard deviation of the lightness for said input image; and

performing a tone correction by using said weighted average value of said average lightness and said standard deviation of the lightness for said input image.

15. A method for specifying a noticeable portion in an input image, said method comprising the steps of:

dividing said input image into a plurality of regions;

with respect to each of said plurality of regions, calculating average lightness, a rate of a sky pixel, which is a pixel satisfying a previously set sky condition, and a rate of a cloud pixel, which is a pixel satisfying a previously set cloud condition, by counting the sky pixel and the cloud pixel, and calculating an average value and a standard deviation of said average lightness for said plurality of regions;

judging whether said input image is in a backlight state on the basis of at least one of said average lightness, said average value and said standard deviation of said average lightness, and said rate of said sky pixel and said rate of said cloud pixel; and

if said input image is judged to be in said backlight state, setting an importance level expressing a most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said average lightness and said average value of said average lightness, and a portion which is not the dark portion due to the backlight but is a portion estimated to be something other than said cloud and said sky on the basis of said rate of said sky pixel and said rate of said cloud pixel, and setting an importance level lower than said importance level expressing the most noticeable portion to other portions.

16. A method for performing a color saturation correction for an input image, said method comprising the steps of:

calculating a statistical amount with respect to color saturation of each pixel of said input image; and

calculating a color saturation correction coefficient by using a color saturation correction reference value expressing a color saturation correction tendency of an operator and said statistical amount with respect to said color saturation, and performing the color saturation correction by said color saturation correction coefficient.

17. The method as set forth in claim 16, further comprising a step of calculating a statistical amount with respect to hue of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a range of the hue by a predetermined number as a hue reference value,

wherein said second calculating step comprises a step of adjusting said color saturation correction coefficient on the basis of said hue reference value.

18. The method as set forth in claim 17, wherein, in said adjusting step, said color saturation correction coefficient is adjusted in accordance with a sine value of a difference between a hue value of each pixel of said input image and said hue reference value.

19. The method as set forth in claim 17, wherein

each of said plurality of hue regions is divided into a first region of a high lightness region and a second region of a low lightness region on the basis of a lightness value of a color having maximum color saturation in each said hue region, and

said reference hue regions are a region among said first regions of all said hue regions, in which most pixels of said input image are included, and a region among said second regions of all said hue regions, in which most pixels of the input image are included.

20. The method as set forth in claim 16, wherein said color saturation correction reference value is a value obtained by averaging differences between an average of color saturation of an image after a color saturation correction by the operator and an average of color saturation of the image before the color saturation correction over a plurality of images.

21. A method for performing a contour emphasis correction for an input image, said method comprising the steps of:

generating a smoothed image by performing a smoothing processing to a process image, which is comprised of lightness components of said input image;

generating a difference image by calculating a difference between said process image and said smoothed image;

calculating a statistical amount with respect to pixel values of said difference image;

calculating a contour emphasis correction coefficient on the basis of a contour emphasis correction reference value expressing a contour emphasis correction tendency of an operator and said statistical amount with respect to said pixel values of said difference image; and

by correcting each pixel value of said difference image by said contour emphasis correction coefficient, and adding the corrected pixel value to a corresponding pixel value of said process image, generating an output image.

22. The method as set forth in claim 21, further comprising a step of generating an image in which the lightness component of each pixel of said input image is made a pixel value of a corresponding pixel of said output image.

23. The method as set forth in claim 21, wherein said contour emphasis correction reference value is a value obtained by averaging averages with respect to an absolute value of a difference between lightness of each pixel of an image after a contour emphasis correction by an operator and lightness of a corresponding pixel of the image before the contour emphasis correction over a plurality of images.

24. The method as set forth in claim 21, wherein, in said step of calculating the contour emphasis correction coefficient, said contour emphasis correction coefficient is calculated by dividing said contour emphasis correction reference value by an average of absolute values of the pixel values, which is a statistical amount with respect to the pixel values of said difference image.

25. A method for performing an image processing, said method comprising the steps of:

performing a color balance correction by using a color fog hue angle component of a color saturation correction reference value, which is calculated by weighting color saturation of each pixel in accordance with a possibility of color fog of each pixel to generate a color fog removal image;

converting values of respective color components of a highlight pixel having highest lightness in said color fog removal image in accordance with a specified maximum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said highlight pixel and a minimum value which the respective color components can take, converting values of respective color components of a shadow pixel having lowest lightness in said color fog removal image in accordance with a specified minimum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said shadow pixel and a maximum value which the respective color components can take, performing a linear conversion with respect to each pixel other than said highlight pixel and said shadow pixel on the basis of the conversion result of said highlight pixel and said shadow pixel to generate a post-range conversion image;

dividing said post-range conversion image into a plurality of regions, calculating, with respect to all of said plurality of regions and each of said plurality of regions, a statistical amount of a human skin pixel satisfying a previously set human skin condition, a statistical amount of a sky pixel, which is a pixel satisfying a previously set sky condition, a statistical amount of a cloud pixel, which is a pixel satisfying a previously set cloud condition, and a statistical amount of lightness, if it is judged that there is a region including a portion estimated to be a person on the basis of said statistical amount of said human skin pixel, setting an importance level expressing a most noticeable portion to said region including the portion estimated to be a person, and setting an importance level lower than said importance level expressing said most noticeable portion to a region including a portion estimated something other than the person on the basis of said statistical amount of said human skin pixel, if it is judged that said post-range conversion image is in a backlight state, setting said importance level expressing said most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said statistical amount of lightness or a portion which is not the dark portion due to the backlight but is estimated to be something other than the cloud and the sky on the basis of said statistical amount of said cloud pixel and said statistical amount of said sky pixel, and setting an importance level lower than said importance level expressing said most noticeable portion to other regions;

by weighting an average lightness and a standard deviation of the lightness calculated with respect to each of said plurality of regions by said importance level set to each of said plurality of regions to calculate a weighted average value of said average lightness and a weighted standard deviation of the lightness with respect to said post-range conversion image, performing a tone correction by using the weighted average value of said average lightness and the weighted standard deviation of the lightness to generate a post-tone correction image;

calculating a statistical amount with respect to color saturation of each pixel of said post-tone correction image, calculating a color saturation correction coefficient by using a color saturation correction reference value expressing a color saturation correction tendency of an operator and said statistical amount with respect to the color saturation, and performing a color saturation correction by said color saturation correction coefficient to generate a post-color saturation correction image; and

generating a smoothed image by performing a smoothing processing to a process image, which is composed of lightness components of said post-color saturation correction image, generating a difference image by calculating a difference between said process image and said smoothed image, calculating a contour emphasis correction coefficient on the basis of a contour emphasis correction reference value expressing a contour emphasis correction tendency of an operator and a statistical amount with respect to pixel values of said difference image, correcting each pixel value of said difference image by said contour emphasis correction coefficient, generating an output image by adding said corrected pixel value to a corresponding pixel value of said process image, and generating an image in which said lightness component of each pixel of said post-color saturation correction image is replaced with a pixel value of a corresponding pixel of said output image.

26. A program embodied on a medium, for causing a computer to correct a color fog for an input image, said program comprising the steps of:

calculating a statistical amount with respect to color saturation components of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a hue given as a correction reference value of said color fog, said color saturation being weighted by values of lightness components of said at least one group of pixels; and

correcting said color fog by using said correction reference value with respect to each pixel of said input image.

27. The program as set forth in claim 26, wherein said reference hue region is a hue region in which most pixels of said input image are included.

28. The method as set forth in claim 26, wherein

each of said plurality of hue regions is divided into a first region of a high lightness region and a second region of a low lightness region on the basis of a lightness value of a color having maximum color saturation in each said hue region, and

said reference hue regions are a region among said first regions of all said hue regions, in which most pixels of said input image are included, and a region among said second regions of all said hue regions, in which most pixels of the input image are included.

29. The program as set forth in claim 26, further comprising a step of calculating a statistical amount with respect to hue of said at least one group of pixels included in said reference hue region as a hue reference value of said color fog, and

said correcting step comprises a step of adjusting said correction reference value on the basis of said hue reference value.

30. The program as set forth in claim 29, wherein, in said adjusting step, said correction reference value is adjusted in accordance with a cosine value of a difference between a hue value of each pixel of said input image and said hue reference value.

31. The program as set forth in claim 26, wherein the weighting is performed such that as a difference with a lightness value of a color having maximum color saturation in said reference hue region becomes large, the weight becomes heavy.

32. A program embodied on a medium, for causing a computer to correct a color fog for an input image, said program comprising the steps of:

calculating a statistical amount with respect to hue values of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a range of hue as a hue reference value of said color fog;

calculating a statistical amount with respect to color saturation components of said at least one group of pixels included in said reference hue region in accordance with a predetermined method as a correction reference value of said color fog; and

correcting said color fog with respect to each pixel of said input image by using said correction reference value adjusted by using said hue reference value.

33. The program as set forth in claim 32, wherein, in said correcting step, said correction reference value is adjusted in accordance with a cosine value of a difference between a hue value of each pixel of said input image and said hue reference value.

34. A program embodied on a medium, for causing a computer to perform a range correction for an input image, said program comprising the steps of:

detecting a highlight pixel having highest lightness and a shadow pixel having lowest lightness in pixels included in said input image;

converting values of respective color components of said highlight pixel in accordance with a specified maximum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said highlight pixel and minimum values which said respective color components can take;

converting values of respective color components of said shadow pixel in accordance with a specified minimum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said shadow pixel and maximum values which said respective color components can take; and

with respect to each of said color components, linearly converting a value of the color component of each pixel of said input image, which is contained in a range from a value of the color component for said shadow pixel before conversion to a value of the color component for said highlight pixel before conversion, into a value of the color component in a range from a value of the color component for said shadow pixel after conversion to a value of the color component for said highlight pixel after conversion.

35. The program as set forth in claim 34, wherein, in said step of converting values of respective color components of said highlight pixel, a maximum value among values of the color components of said highlight pixel is converted to the specified maximum gradation value, and

in said step of converting values of respective color components of said shadow pixel, a minimum value among values of the color components of said shadow pixel is converted to the specified minimum gradation value.

36. A program embodied on a medium, for causing a computer to specify a noticeable portion in an input image, said program comprising the steps of:

dividing said input image into a plurality of regions;

calculating a rate of a human skin pixel by counting said human skin pixel, which is a pixel satisfying a previously set human skin condition, with respect to each of said plurality of regions, and calculating an average and a standard deviation of said rate of said human skin pixel for said plurality of regions;

judging whether or not a region including a portion estimated to be a person on the basis of said average of said rate of said human skin pixel exists; and

if it is judged that the region including the portion estimated to be the person exists, setting an importance level expressing a most noticeable portion to the region including the portion estimated to be the person, and setting an importance level lower than said importance level expressing the most noticeable portion to a region including a portion estimated to be something other than the person on the basis of a value of the standard deviation.

37. The program as set forth in claim 36, further comprising the steps of:

with respect to each of said plurality of regions, calculating average lightness, a rate of a sky pixel, which is a pixel satisfying a previously set sky condition, and a rate of a cloud pixel, which is a pixel satisfying a previously set cloud condition, by counting a number of said sky pixel and a number of said cloud pixel, and calculating an average value and a standard deviation of said average lightness for the plurality of regions;

judging whether said input image is in a backlight state on the basis of at least one of said average lightness, said average value and said standard deviation of said average lightness, and said rate of said sky pixel and said rate of said cloud pixel; and

if said input image is judged to be in said backlight state, setting an importance level expressing a most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said average lightness and said average value of said average lightness, or a portion which is not the dark portion due to the backlight but is a portion estimated to be something other than the cloud and the sky on the basis of said rate of said sky pixel and said rate of said cloud pixel, and setting an importance level lower than said importance level expressing the most noticeable portion to other regions.

38. The program as set forth in claim 37, further comprising the steps of:

if said input image is not judged to be an image including the region including the portion estimated to be the person and an image in said backlight state, setting the importance level expressing the most noticeable portion to the respective regions.

39. The program as set forth in claim 38, further comprising the steps of:

by weighting said average lightness and said standard deviation of the lightness, which are calculated with respect to each of said plurality of regions, by the importance level set for each of said plurality of regions, calculating a weighted average value of said average lightness and a weighted standard deviation of the lightness for said input image; and

performing a tone correction by using said weighted average value of said average lightness and said standard deviation of the lightness for said input image.

40. A program embodied on a medium, for causing a computer to specify a noticeable portion in an input image, said program comprising the steps of:

dividing said input image into a plurality of regions;

with respect to each of said plurality of regions, calculating average lightness, a rate of a sky pixel, which is a pixel satisfying a previously set sky condition, and a rate of a cloud pixel, which is a pixel satisfying a previously set cloud condition, by counting the sky pixel and the cloud pixel, and calculating an average value and a standard deviation of said average lightness for said plurality of regions;

judging whether said input image is in a backlight state on the basis of at least one of said average lightness, said average value and said standard deviation of said average lightness, and said rate of said sky pixel and said rate of said cloud pixel; and

if said input image is judged to be in said backlight state, setting an importance level expressing a most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said average lightness and said average value of said average lightness, and a portion which is not the dark portion due to the backlight but is a portion estimated to be something other than said cloud and said sky on the basis of said rate of said sky pixel and said rate of said cloud pixel, and setting an importance level lower than said importance level expressing the most noticeable portion to other portions.

41. A program embodied on a medium, for causing a computer to perform a color saturation correction for an input image, said program comprising the steps of:

calculating a statistical amount with respect to color saturation of each pixel of said input image; and

calculating a color saturation correction coefficient by using a color saturation correction reference value expressing a color saturation correction tendency of an operator and said statistical amount with respect to said color saturation, and performing the color saturation correction by said color saturation correction coefficient.

42. The program as set forth in claim 41, further comprising a step of calculating a statistical amount with respect to hue of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a range of the hue by a predetermined number as a hue reference value,

wherein said second calculating step comprises a step of adjusting said color saturation correction coefficient on the basis of said hue reference value.

43. The program as set forth in claim 42, wherein, in said adjusting step, said color saturation correction coefficient is adjusted in accordance with a sine value of a difference between a hue value of each pixel of said input image and said hue reference value.

44. The program as set forth in claim 42, wherein

each of said plurality of hue regions is divided into a first region of a high lightness region and a second region of a low lightness region on the basis of a lightness value of a color having maximum color saturation in each said hue region, and

said reference hue regions are a region among said first regions of all said hue regions, in which most pixels of said input image are included, and a region among said second regions of all said hue regions, in which most pixels of the input image are included.

45. The program as set forth in claim 41, wherein said color saturation correction reference value is a value obtained by averaging differences between an average of color saturation of an image after a color saturation correction by the operator and an average of color saturation of the image before the color saturation correction over a plurality of images.

46. A program embodied on a medium, for causing a computer to perform a contour emphasis correction for an input image, said program comprising the steps of:

generating a smoothed image by performing a smoothing processing to a process image, which is comprised of lightness components of said input image;

generating a difference image by calculating a difference between said process image and said smoothed image;

calculating a statistical amount with respect to pixel values of said difference image;

calculating a contour emphasis correction coefficient on the basis of a contour emphasis correction reference value expressing a contour emphasis correction tendency of an operator and said statistical amount with respect to said pixel values of said difference image; and

by correcting each pixel value of said difference image by said contour emphasis correction coefficient, and adding the corrected pixel value to a corresponding pixel value of said process image, generating an output image.

47. The program as set forth in claim 46, further comprising a step of generating an image in which the lightness component of each pixel of said input image is made a pixel value of a corresponding pixel of said output image.

48. The program as set forth in claim 46, wherein said contour emphasis correction reference value is a value obtained by averaging averages with respect to an absolute value of a difference between lightness of each pixel of an image after a contour emphasis correction by an operator and lightness of a corresponding pixel of the image before the contour emphasis correction over a plurality of images.

49. The program as set forth in claim 46, wherein, in said step of calculating the contour emphasis correction coefficient, said contour emphasis correction coefficient is calculated by dividing said contour emphasis correction reference value by an average of absolute values of the pixel values, which is a statistical amount with respect to the pixel values of said difference image.

50. A program embodied on a medium, for causing a computer to perform an image processing, said program comprising the steps of:

performing a color balance correction by using a color fog hue angle component of a color saturation correction reference value, which is calculated by weighting color saturation of each pixel in accordance with a possibility of color fog of each pixel to generate a color fog removal image;

converting values of respective color components of a highlight pixel having highest lightness in said color fog removal image in accordance with a specified maximum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said highlight pixel and a minimum value which the respective color components can take, converting values of respective color components of a shadow pixel having lowest lightness in said color fog removal image in accordance with a specified minimum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said shadow pixel and a maximum value which the respective color components can take, performing a linear conversion with respect to each pixel other than said highlight pixel and said shadow pixel on the basis of the conversion result of said highlight pixel and said shadow pixel to generate a post-range conversion image;

dividing said post-range conversion image into a plurality of regions, calculating, with respect to all of said plurality of regions and each of said plurality of regions, a statistical amount of a human skin pixel satisfying a previously set human skin condition, a statistical amount of a sky pixel, which is a pixel satisfying a previously set sky condition, a statistical amount of a cloud pixel, which is a pixel satisfying a previously set cloud condition, and a statistical amount of lightness, if it is judged that there is a region including a portion estimated to be a person on the basis of said statistical amount of said human skin pixel, setting an importance level expressing a most noticeable portion to said region including the portion estimated to be a person, and setting an importance level lower than said importance level expressing said most noticeable portion to a region including a portion estimated something other than the person on the basis of said statistical amount of said human skin pixel, if it is judged that said post-range conversion image is in a backlight state, setting said importance level expressing said most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said statistical amount of lightness or a portion which is not the dark portion due to the backlight but is estimated to be something other than the cloud and the sky on the basis of said statistical amount of said cloud pixel and said statistical amount of said sky pixel, and setting an importance level lower than said importance level expressing said most noticeable portion to other regions;

by weighting an average lightness and a standard deviation of the lightness calculated with respect to each of said plurality of regions by said importance level set to each of said plurality of regions to calculate a weighted average value of said average lightness and a weighted standard deviation of the lightness with respect to said post-range conversion image, performing a tone correction by using the weighted average value of said average lightness and the weighted standard deviation of the lightness to generate a post-tone correction image;

calculating a statistical amount with respect to color saturation of each pixel of said post-tone correction image, calculating a color saturation correction coefficient by using a color saturation correction reference value expressing a color saturation correction tendency of an operator and said statistical amount with respect to the color saturation, and performing a color saturation correction by said color saturation correction coefficient to generate a post-color saturation correction image; and

generating a smoothed image by performing a smoothing processing to a process image, which is composed of lightness components of said post-color saturation correction image, generating a difference image by calculating a difference between said process image and said smoothed image, calculating a contour emphasis correction coefficient on the basis of a contour emphasis correction reference value expressing a contour emphasis correction tendency of an operator and a statistical amount with respect to pixel values of said difference image, correcting each pixel value of said difference image by said contour emphasis correction coefficient, generating an output image by adding said corrected pixel value to a corresponding pixel value of said process image, and generating an image in which said lightness component of each pixel of said post-color saturation correction image is replaced with a pixel value of a corresponding pixel of said output image.

51. An apparatus for correcting a color fog for an input image, comprising:

a calculator for calculating a statistical amount with respect to color saturation components of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a hue given as a correction reference value of said color fog, said color saturation being weighted by values of lightness components of said at least one group of pixels; and

means for correcting said color fog by using said correction reference value with respect to each pixel of said input image.

52. The apparatus as set forth in claim 51, wherein said reference hue region is a hue region in which most pixels of said input image are included.

53. The apparatus as set forth in claim 51, wherein

each of said plurality of hue regions is divided into a first region of a high lightness region and a second region of a low lightness region on the basis of a lightness value of a color having maximum color saturation in each said hue region, and

said reference hue regions are a region among said first regions of all said hue regions, in which most pixels of said input image are included, and a region among said second regions of all said hue regions, in which most pixels of the input image are included.

54. The apparatus as set forth in claim 51, further comprising a second calculator for calculating a statistical amount with respect to hue of said at least one group of pixels included in said reference hue region as a hue reference value of said color fog, and

said means for correcting comprises an adjuster for adjusting said correction reference value on the basis of said hue reference value.

55. The apparatus as set forth in claim 54, wherein said means for adjusting adjusts said correction reference value in accordance with a cosine value of a difference between a hue value of each pixel of said input image and said hue reference value.

56. The apparatus as set forth in claim 51, wherein the weighting is performed such that as a difference with a lightness value of a color having maximum color saturation in said reference hue region becomes large, the weight becomes heavy.

57. An apparatus for correcting a color fog for an input image, comprising:

a first calculator for calculating a statistical amount with respect to hue values of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a range of hue as a hue reference value of said color fog;

a second calculator for calculating a statistical amount with respect to color saturation components of said at least one group of pixels included in said reference hue region in accordance with a predetermined method as a correction reference value of said color fog; and

means for correcting said color fog with respect to each pixel of said input image by using said correction reference value adjusted by using said hue reference value.

58. The apparatus as set forth in claim 57, wherein said means for correcting adjusts said correction reference value in accordance with a cosine value of a difference between a hue value of each pixel of said input image and said hue reference value.

59. An apparatus for performing a range correction for an input image, comprising:

a detector for detecting a highlight pixel having highest lightness and a shadow pixel having lowest lightness in pixels included in said input image;

a first converter for converting values of respective color components of said highlight pixel in accordance with a specified maximum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said highlight pixel and minimum values which said respective color components can take;

a second converter for converting values of respective color components of said shadow pixel in accordance with a specified minimum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said shadow pixel and maximum values which said respective color components can take; and

a third converter for linearly converting, with respect to each of said color components, a value of the color component of each pixel of said input image, which is contained in a range from a value of the color component for said shadow pixel before conversion to a value of the color component for said highlight pixel before conversion, into a value of the color component in a range from a value of the color component for said shadow pixel after conversion to a value of the color component for said highlight pixel after conversion.

60. The apparatus as set forth in claim 59, wherein said first converter converts a maximum value among values of the color components of said highlight pixel to the specified maximum gradation value, and

said second converter converts a minimum value among values of the color components of said shadow pixel to the specified minimum gradation value.

61. An apparatus for specifying a noticeable portion in an input image, comprising:

a divider for dividing said input image into a plurality of regions;

a calculator for calculating a rate of a human skin pixel by counting said human skin pixel, which is a pixel satisfying a previously set human skin condition, with respect to each of said plurality of regions, and calculating an average and a standard deviation of said rate of said human skin pixel for said plurality of regions;

means for judging whether or not a region including a portion estimated to be a person on the basis of said average of said rate of said human skin pixel exists; and

means for setting, in case where it is judged that the region including the portion estimated to be the person exists, an importance level expressing a most noticeable portion to the region including the portion estimated to be the person, and setting an importance level lower than said importance level expressing the most noticeable portion to a region including a portion estimated to be something other than the person on the basis of a value of the standard deviation.

62. The apparatus as set forth in claim 61, comprising:

a second calculator for calculating, with respect to each of said plurality of regions, average lightness, a rate of a sky pixel, which is a pixel satisfying a previously set sky condition, and a rate of a cloud pixel, which is a pixel satisfying a previously set cloud condition, by counting a number of said sky pixel and a number of said cloud pixel, and calculating an average value and a standard deviation of said average lightness for the plurality of regions;

means for judging whether said input image is in a backlight state on the basis of at least one of said average lightness, said average value and said standard deviation of said average lightness, and said rate of said sky pixel and said rate of said cloud pixel; and

means for setting, in case where said input image is judged to be in said backlight state, an importance level expressing a most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said average lightness and said average value of said average lightness, or a portion which is not the dark portion due to the backlight but is a portion estimated to be something other than the cloud and the sky on the basis of said rate of said sky pixel and said rate of said cloud pixel, and setting an importance level lower than said importance level expressing the most noticeable portion to other regions.

63. The apparatus as set forth in claim 62, further comprising:

means for setting, in case where said input image is not judged to be an image including the region including the portion estimated to be the person and an image in said backlight state, the importance level expressing the most noticeable portion to the respective regions.

64. The apparatus as set forth in claim 63, further comprising:

means for weighting said average lightness and said standard deviation of the lightness, which are calculated with respect to each of said plurality of regions, by the importance level set for each of said plurality of regions, and calculating a weighted average value of said average lightness and a weighted standard deviation of the lightness for said input image; and

means for performing a tone correction by using said weighted average value of said average lightness and said standard deviation of the lightness for said input image.

65. An apparatus for specifying a noticeable portion in an input image, comprising:

a divider for dividing said input image into a plurality of regions;

a calculator for calculating, with respect to each of said plurality of regions, average lightness, a rate of a sky pixel, which is a pixel satisfying a previously set sky condition, and a rate of a cloud pixel, which is a pixel satisfying a previously set cloud condition, by counting the sky pixel and the cloud pixel, and calculating an average value and a standard deviation of said average lightness for said plurality of regions;

means for judging whether said input image is in a backlight state on the basis of at least one of said average lightness, said average value and said standard deviation of said average lightness, and said rate of said sky pixel and said rate of said cloud pixel; and

means for setting, in case where said input image is judged to be in said backlight state, an importance level expressing a most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said average lightness and said average value of said average lightness, and a portion which is not the dark portion due to the backlight but is a portion estimated to be something other than said cloud and said sky on the basis of said rate of said sky pixel and said rate of said cloud pixel, and setting an importance level lower than said importance level expressing the most noticeable portion to other portions.

66. An apparatus for performing a color saturation correction for an input image, comprising:

a first calculator for calculating a statistical amount with respect to color saturation of each pixel of said input image; and

a second calculator for calculating a color saturation correction coefficient by using a color saturation correction reference value expressing a color saturation correction tendency of an operator and said statistical amount with respect to said color saturation, and performing the color saturation correction by said color saturation correction coefficient.

67. The apparatus as set forth in claim 66, further comprising a third calculator for calculating a statistical amount with respect to hue of at least one group of pixels included in a reference hue region of a plurality of hue regions generated by dividing a range of the hue by a predetermined number as a hue reference value,

wherein said second calculator comprises an adjuster for adjusting said color saturation correction coefficient on the basis of said hue reference value.

68. The apparatus as set forth in claim 67, wherein said adjuster adjusts said color saturation correction coefficient in accordance with a sine value of a difference between a hue value of each pixel of said input image and said hue reference value.

69. The apparatus as set forth in claim 67, wherein

each of said plurality of hue regions is divided into a first region of a high lightness region and a second region of a low lightness region on the basis of a lightness value of a color having maximum color saturation in each said hue region, and

said reference hue regions are a region among said first regions of all said hue regions, in which most pixels of said input image are included, and a region among said second regions of all said hue regions, in which most pixels of the input image are included.

70. The apparatus as set forth in claim 66, wherein said color saturation correction reference value is a value obtained by averaging differences between an average of color saturation of an image after a color saturation correction by the operator and an average of color saturation of the image before the color saturation correction over a plurality of images.

71. An apparatus for performing a contour emphasis correction for an input image, comprising:

a first generator for generating a smoothed image by performing a smoothing processing to a process image, which is comprised of lightness components of said input image;

a second generator for generating a difference image by calculating a difference between said process image and said smoothed image;

a first calculator for calculating a statistical amount with respect to pixel values of said difference image;

a second calculator for calculating a contour emphasis correction coefficient on the basis of a contour emphasis correction reference value expressing a contour emphasis correction tendency of an operator and said statistical amount with respect to said pixel values of said difference image; and

means for correcting each pixel value of said difference image by said contour emphasis correction coefficient, and adding the corrected pixel value to a corresponding pixel value of said process image to generate an output image.

72. The apparatus as set forth in claim 71, further comprising a third generator for generating an image in which the lightness component of each pixel of said input image is made a pixel value of a corresponding pixel of said output image.

73. The apparatus as set forth in claim 71, wherein said contour emphasis correction reference value is a value obtained by averaging averages with respect to an absolute value of a difference between lightness of each pixel of an image after a contour emphasis correction by an operator and lightness of a corresponding pixel of the image before the contour emphasis correction over a plurality of images.

74. The apparatus as set forth in claim 71, wherein said means for correcting calculates said contour emphasis correction coefficient by dividing said contour emphasis correction reference value by an average of absolute values of the pixel values, which is a statistical amount with respect to the pixel values of said difference image.

75. An apparatus for performing an image processing, comprising:

means for performing a color balance correction by using a color fog hue angle component of a color saturation correction reference value, which is calculated by weighting color saturation of each pixel in accordance with a possibility of color fog of each pixel to generate a color fog removal image;

means for converting values of respective color components of a highlight pixel having highest lightness in said color fog removal image in accordance with a specified maximum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said highlight pixel and a minimum value which the respective color components can take, converting values of respective color components of a shadow pixel having lowest lightness in said color fog removal image in accordance with a specified minimum gradation value so as not to change a ratio of gradation differences between values of the respective color components of said shadow pixel and a maximum value which the respective color components can take, performing a linear conversion with respect to each pixel other than said highlight pixel and said shadow pixel on the basis of the conversion result of said highlight pixel and said shadow pixel to generate a post-range conversion image;

means for dividing said post-range conversion image into a plurality of regions, calculating, with respect to all of said plurality of regions and each of said plurality of regions, a statistical amount of a human skin pixel satisfying a previously set human skin condition, a statistical amount of a sky pixel, which is a pixel satisfying a previously set sky condition, a statistical amount of a cloud pixel, which is a pixel satisfying a previously set cloud condition, and a statistical amount of lightness, if it is judged that there is a region including a portion estimated to be a person on the basis of said statistical amount of said human skin pixel, setting an importance level expressing a most noticeable portion to said region including the portion estimated to be a person, and setting an importance level lower than said importance level expressing said most noticeable portion to a region including a portion estimated something other than the person on the basis of said statistical amount of said human skin pixel, if it is judged that said post-range conversion image is in a backlight state, setting said importance level expressing said most noticeable portion to a region including a portion estimated to be a dark portion due to backlight on the basis of said statistical amount of lightness or a portion which is not the dark portion due to the backlight but is estimated to be something other than the cloud and the sky on the basis of said statistical amount of said cloud pixel and said statistical amount of said sky pixel, and setting an importance level lower than said importance level expressing said most noticeable portion to other regions;

means for weighting an average lightness and a standard deviation of the lightness calculated with respect to each of said plurality of regions by said importance level set to each of said plurality of regions to calculate a weighted average value of said average lightness and a weighted standard deviation of the lightness with respect to said post-range conversion image, and performing a tone correction by using the weighted average value of said average lightness and the weighted standard deviation of the lightness to generate a post-tone correction image;

means for calculating a statistical amount with respect to color saturation of each pixel of said post-tone correction image, calculating a color saturation correction coefficient by using a color saturation correction reference value expressing a color saturation correction tendency of an operator and said statistical amount with respect to the color saturation, and performing a color saturation correction by said color saturation correction coefficient to generate a post-color saturation correction image; and

means for generating a smoothed image by performing a smoothing processing to a process image, which is composed of lightness components of said post-color saturation correction image, generating a difference image-by calculating a difference between said process image and said smoothed image, calculating a contour emphasis correction coefficient on the basis of a contour emphasis correction reference value expressing a contour emphasis correction tendency of an operator and a statistical amount with respect to pixel values of said difference image, correcting each pixel value of said difference image by said contour emphasis correction coefficient, generating an output image by adding said corrected pixel value to a corresponding pixel value of said process image, and generating an image in which said lightness component of each pixel of said post-color saturation correction image is replaced with a pixel value of a corresponding pixel of said output image.